FEATURE — Spring 2008

Professor Jonathan Turner is known as an exemplary teacher, organizing complex material in a way that is understandable to students. Here, he works with graduate student Charles Wiseman, discussing an ATM Switch Card.

Reinventing the Internet

Jonathan Turner, professor of computer science, is helping create a future where virtual worlds would allow people to collaborate internationally and, in the process, help solve global energy and environmental problems.

By Rick Skwiot

When Jonathan Turner gazes into the future, he sees the world’s intractable energy, environmental, and socio-political problems solved in part by algorithms and high-speed Internet routers—an Internet that operates with technology he helped develop and that he is working to reinvent.

His plan to ameliorate the world’s energy ills To move electronic bits instead of people and replace eight-lane highways with fiber-optic cable. That is, to use a new, more powerful and dynamic Internet to give people worldwide virtual presence virtually anywhere in the world.

“There’s almost no limit to the capabilities of programmable routers versus today’s routers, which are fixed. Programmable routers allow virtual presence applications that are a crucial component to address the world’s big challenges,” says Turner, the Barbara J. and Jerome R. Cox, Jr. Professor of Computer Science in the School of Engineering, whose pivotal work over the last 25 years has helped enable today’s Internet. That work includes helping develop technology for high-speed routers—the electronic devices that direct data across a digital network and that form the backbone of today’s Internet.

Today’s global challenges are often tied to questions of energy and the environment, says Turner, who also serves as chair of the Department of Computer Science & Engineering.

“Developing nations’ desire to improve their standards of living juxtaposed with global environmental constraints creates significant challenges that must be addressed,” Turner says. The problems are compounded when coupled with “peaking oil production that may soon make it impossible to meet growing demands.”

But the need for travel and oil consumption can be reduced greatly by allowing virtual presence, says Turner, “that is the ability for a person to be somewhere else and interact in virtual environments that have real-world impact, to give people compelling alternatives to travel.”

He says that the popular virtual Internet world “Second Life”—where people assume the guise of digital characters called avatars, then interact and, in some cases, even do actual business—“allows you to see the potential. It will allow virtual conferences where new technology will enable avatars to simulate your movement and expression through audio and video capture of the user. We can create virtual worlds to let people collaborate across a global environment.”

This approximate face-to-face communication could supplant significant business, educational, and other travel, according to Turner. It even would allow a writer, say, to interview a computer science professor from a remote location and see his body language, gestures, and expressions, Turner contends.

Reinvigorating the Internet
Over the past few years, Turner has been playing a leadership role in the U.S. government’s National Science Foundation (NSF) project GENI (Global Environment for Networking Innovations), which is exploring ways to significantly advance—or reinvent—a 21st-century Internet. As principal investigator on numerous NSF research grants, Turner has been working to develop and deploy future Internet architecture based on network virtualization, or diversification.

Over the past few years, Turner has been playing a leadership role in the U.S. government’s National Science Foundation (NSF) project GENI (Global Environment for Networking Innovations), which is exploring ways to significantly advance—or reinvent—a 21st-century Internet.

“We’re designing systems to overcome the limitations of the current Internet. Today’s network servers are not very good at sophisticated voice and video,” says Turner. “For a decade we’ve had technical solutions, but they are hard to implement because the Internet is so fragmented, with 10,000 service providers, all of whom would have to agree” on a new technology. “It’s hard to get 10 people—much less 10,000—to agree on anything,” he says.
Turner’s interim solution to that dilemma is the development of “overlay” networks—enhanced data-transmission networks with new applications added on top of the current Internet but owned and operated privately and employing advanced routing technology. He is particularly interested in overlay hosting services, in which, he says, “you can build a number of overlays that can be deployed by a user without investing in staff, space, or equipment.”

Algorithms and hardware
Developing such useful and accessible applications for his sophisticated technology is typical of Turner’s approach, says Jerome R. Cox, Jr., senior professor of computer science and former department chairman. “Jon believes engineers have a responsibility to bring their ideas to the public—a responsibility that leads to tangible products in the marketplace.”

Indeed, while much of Turner’s work involves computational algorithms and analysis and theorizing about their potential, he is also focused on developing systems that work today in the real world.

In 1989, along with Cox and colleague Guru M. Parulkar, Turner founded the Applied Research Laboratory (ARL) in Washington University’s School of Engineering to develop advanced computer-science technology. Eight years later, to leverage technology and products they had devised at ARL, the three men formed Growth Networks, Inc., a start-up company that developed advanced switching components for Internet routers and multiservice switching systems. Within two years, they sold the company to Cisco Systems, Inc. for $350 million in Cisco stock. The Growth Networks team then went on to design the switching technology for Cisco’s CRS-1, their flagship router product for carrier networks, building on the technology and ideas created within Growth Networks. Recently, AT&T announced a major commitment to deploy the CRS-1 in their global network, a commitment that could ultimately generate $500 million in sales for Cisco.

Parulkar, now consulting professor and executive director of the Clean Slate Internet Design Program at Stanford University, says Turner’s work will continue to be critical to the evolution of the Internet.

“Jon has one of the strongest records of innovations for the past 25 years in the broad area of networking,” says Parulkar. “He has written a number of seminal research papers, developed and demonstrated key technologies, and also successfully transferred them to industry—very few people in the world have his depth and breadth.”

Parulkar characterizes Turner as a forward thinker whose work will influence how the world communicates in the 21st century.

“A few years ago Jon was one of the very select few who saw the need to think beyond the current Internet and made the case for the urgent and important need to ‘reinvent the Internet,’” says Parulkar. “Jon proposed and articulated the idea of network virtualization to allow multiple, very different network architectures to coexist on a single physical substrate. Though the jury is still out, more and more people are starting to see the potential of network virtualization as the foundation for a ‘Future Internet’ that would help address limitations of the current Internet and also make this ‘Future Internet’ evolvable—very, very important going forward.”

In 2007, Turner was recognized “for contributions to the design and analysis of high-performance communication networks” by being elected to the prestigious National Academy of Engineering. He is one of two engineering faculty elected to the academy.

Serendipitous summer
But Turner’s contributions to computer science might never have come about except for quirks in course scheduling at Washington University over three decades ago. When he transferred here as an undergraduate in the summer of 1975, he thought he would major in civil engineering.

“But for the summer semester there were a limited number of courses offered,” says Turner, and nothing in civil engineering. “So I took computer science and liked it”—even though, he recalls, the state of computer science on campus those days was embodied in an IBM 360 that relied on keypunched data cards and card-sorters to perform computations.

“We used to wire individual logic gates together,” Turner recalls. “Now we have programmable logic devices containing tens of thousands of logic gates.”

His burgeoning interest in computers led to a double major in computer science and electrical engineering and ultimately to a job at Bell Labs in Chicago, where, he says, “I learned about communications networks and really got involved.”

While in Chicago, he earned a master’s and a doctorate in computer science at Northwestern University and returned to Washington University in 1983 as an assistant professor of computer science. As his research and technological achievements grew, so did his ability to explain the sophisticated concepts he developed to students in the classroom.

Says Cox: “Jon has been an exemplary teacher. He demands and gets a quality effort from his students because they are inspired by how he organizes complex material in a manner that is understandable to them.”

Though Turner conducts, as expected, graduate research seminars on sophisticated computer networking, he also teaches Introduction to Digital Logic and Computer Design to first- and second-year undergraduate engineering students.

“We’re limited more by our imaginations than by our technology,” says Turner, who received the Arthur Holly Compton Faculty Achievement Award in 2004. “Younger minds are racing ahead and finding new applications. I’m putting tools in the hands of bright young people who think of new ways to use technology that older guys like me never would.”

Rick Skwiot is a freelance writer based in St. Louis.